The effect of carbon on the formation of sigma phase in austenitic stainless steel

An investigation was undertaken to study the effects of carbon content (and chromium content) on the amount and rate of sigma phase formation in experimentally prepared iron-chromium-nickel-carbon alloys. Nine alloys were melted in a laboratory under vacuum or controlled atmosphere using high-purity raw materials as ingredients. The compositions were typically in the range for austenitic stainless steels, with a systematic variation of chromium and carbon content. The composition ranges for each alloying element varied as follows: carbon 0-0.40 wt.%, chromium 18-26 wt.%, nickel 14-17 wt.%, and balance iron.

Prior to aging, the alloys were grouped into three types based upon the pretreatments they received—namely, as-cast, cold-worked, and solution-treated conditions. The aging temperatures varied from 650- 850°C (1200-1560°F), and the aging time ranged from 1-1000 hours.

Techniques such as x-ray diffraction, transmission electron microscopy, scanning electron microscopy, metallographic staining, and Magne Gage measurements (for ferrite content) were used to identify and characterize the phases present in the samples. The weight fraction of sigma phase was determined by quantitative metallographic techniques.

The sigma formation tendencies of the alloys were compared to and analyzed in terms of two semi-empirical theories: 1) the equivalent chromium content theory of Hull⁽¹⁴⁾ and 2) the electron vacancy theory developed by Woodyatt et al.⁽¹¹⁷⁾

The major findings of this investigation include:

1. An increase of carbon content of the austenitic stainless steel alloys in the range of 0-0.40 wt.% retards sigma formation at a given equivalent chromium content.

2. The major source of the effect of carbon was the removal of the carbide (and sigma-forming) elements from solution by precipitation of carbides.

3. The most successful and quantitative analysis etchant was KOH-H₂O solution with controlled current density and etching time. This solution stains sigma phase to a reddish brown color, whereas other popular etchants, such as oxalic acid, HC1-HNO₃-H₂0 solution, etc., do not.

4. Optical metallography was more suitable to estimate the percent sigma in the matrix than the bulk electrolytic extraction method because of suspected dissolution of sigma in the 10% HCl-90% CH₃OH solution as found by our studies, as well as numerous others reported in the literature.

5. Both the equivalent chromium content (ECC) and the electron vacancy number (Nv̄) theories are useful for predicting the formation of sigma phase in the austenitic stainless steels. Alloys with overall ECC > 17 ± 0.05 or Nv̄ of 2.85 ± 0.05 form sigma in the austenitic stainless steels. This value agrees well with a minimum Nv̄ of 2.95-3.05 for sigma formation in Fe-Cr-Co-Ni-Mo alloys reported by de Barbadillo in the 1977 Handbook of Stainless Steels.

6. The as-cast or cold-worked structures form sigma at a greater rate than the solution-treated alloys.

7. Ferrite in the austenitic stainless steel promotes sigma formation by the nucleation of sigma within the ferrite or at the austenite-ferrite interface.

8. The rate of sigma formation was greater at 750°C (1380°F) than at either 650°C (1200°F) or 850°C (1560'°F).

9. The morphology of sigma phase varied. It was semicontinuous with massive dog-bone shape in the as-cast alloys to needle-like or massive globular (shaped) in the cold-worked or solution-treated alloys.